| 1. |
- Wirén, Mats, 1954-, et al.
(author)
-
Annotating the Narrative: A Plot of Scenes, Events, Characters and Other Intriguing Elements
- 2022
-
In: LIVE and LEARN. - Gothenburg : Department of Swedish, Multilingualism, Language Technology. - 9789187850837 ; , s. 161-164
-
Book chapter (other academic/artistic)abstract
- Analysis of narrative structure in prose fiction is a field which is gaining increased attention in NLP, and which potentially has many interesting and more far-reaching applications. This paper provides a summary and motivation of two different but interrelated strands of work that we have carried out in this field during the last years: on the one hand, principles and guidelines for annotation, and on the other, methods for automatic annotation.
|
|
| 2. |
- Amanaki, Erini, et al.
(author)
-
Fine-grained Entailment: Resources for Greek NLI and Precise Entailment
- 2022
-
In: Proceedings of the Workshop on Dataset Creation for Lower-Resourced Languages within the 13th Language Resources and Evaluation Conference. - Marseille, France : European Language Resources Association (ELRA). - 9782493814067
-
Conference paper (peer-reviewed)abstract
- In this paper, we present a number of fine-grained resources for Natural Language Inference (NLI). In particular, we present a number of resources and validation methods for Greek NLI and a resource for precise NLI. First, we extend the Greek version of the FraCaS test suite to include examples where the inference is directly linked to the syntactic/morphological properties of Greek. The new resource contains an additional 428 examples, making it in total a dataset of 774 examples. Expert annotators have been used in order to create the additional resource, while extensive validation of the original Greek version of the FraCaS by non-expert and expert subjects is performed. Next, we continue the work initiated by (CITATION), according to which a subset of the RTE problems have been labeled for missing hypotheses and we present a dataset an order of magnitude larger, annotating the whole SuperGlUE/RTE dataset with missing hypotheses. Lastly, we provide a de-dropped version of the Greek XNLI dataset, where the pronouns that are missing due to the pro-drop nature of the language are inserted. We then run some models to see the effect of that insertion and report the results.
|
|
| 3. |
- Bernardy, Jean-Philippe, 1978, et al.
(author)
-
Can the Transformer Learn Nested Recursion with Symbol Masking?
- 2021
-
In: Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021, August 1 - 6, 2021, Online / Chengqing Zong, Fei Xia, Wenjie Li, Roberto Navigli (Editors). - Stroudsburg, PA : The Association for Computational Linguistics. - 9781954085541
-
Conference paper (peer-reviewed)abstract
- We investigate if, given a simple symbol masking strategy, self-attention models are capable of learning nested structures and generalise over their depth. We do so in the simplest setting possible, namely languages consisting of nested parentheses of several kinds. We use encoder-only models, which we train to predict randomly masked symbols, in a BERT-like fashion. We find that the accuracy is well above random baseline, with accuracy consistently above 50% both when increasing nesting depth and distances between training and testing. However, we find that the predictions made correspond to a simple parenthesis counting strategy, rather than a push-down automaton. This suggests that self-attention models are not suitable for tasks which require generalisation to more complex instances of recursive structures than those found in the training set.
|
|
| 4. |
- Dobnik, Simon, 1977, et al.
(author)
-
In Search of Meaning and Its Representations for Computational Linguistics
- 2022
-
In: Proceedings of the 2022 CLASP Conference on (Dis)embodiment, Gothenburg and online 15–16 September 2022 / Simon Dobnik, Julian Grove and Asad Sayeed (eds.). - : Association for Computational Linguistics. - 2002-9764. - 9781955917674
-
Conference paper (peer-reviewed)abstract
- In this paper we examine different meaning representations that are commonly used in different natural language applications today and discuss their limits, both in terms of the aspects of the natural language meaning they are modelling and in terms of the aspects of the application for which they are used.
|
|
| 5. |
|
|
| 6. |
- Ek, Adam, 1990, et al.
(author)
-
Can argument-predicate relationships be extracted from UD trees?
- 2021
-
In: Proceedings of the Joint 15th Linguistic Annotation Workshop (LAW) and 3rd Designing Meaning Representations (DMR) Workshop, November 2021. - Punta Cana, Dominican Republic : Association for Computational Linguistics. - 9781954085855
-
Conference paper (peer-reviewed)abstract
- In this paper we investigate the possibility of extracting predicate-argument relations from UD trees (and enhanced UD graphs). Concretely, we apply UD parsers on an English question answering/semantic-role labeling data set (FitzGerald et al., 2018) and check if the annotations reflect the relations in the resulting parse trees, using a small number of rules to extract this information. We find that 79.1 % of the argument-predicate pairs can be found in this way, on the basis of Udify (Kondratyuk and Straka, 2019). Error analysis reveals that half of the error cases are attributable to shortcomings in the dataset. The remaining errors are mostly due to predicate-argument relations not being extractible algorithmically from the UD trees (requiring semantic reasoning to be resolved). The parser itself is only responsible for a small portion of errors. Our analysis suggests a number of improvements to the UD annotation schema: we propose to enhance the schema in four ways, in order to capture argument-predicate relations. Additionally, we propose improvements regarding data collection for question answering/semantic-role labeling data.
|
|
| 7. |
- Ek, Adam, 1990, et al.
(author)
-
Composing Byte-Pair Encodings for Morphological Sequence Classification
- 2020
-
In: Proceedings of the Fourth Workshop on Universal Dependencies (UDW 2020), December 13, 2020, Barcelona, Spain (Online). - : Association for Computational Linguistics. - 0891-2017. - 9781952148484
-
Conference paper (peer-reviewed)abstract
- Byte-pair encodings is a method for splitting a word into sub-word tokens, a language model then assigns contextual representations separately to each of these tokens. In this paper, we evaluate four different methods of composing such sub-word representations into word representations. We evaluate the methods on morphological sequence classification, the task of predicting grammatical features of a word. Our experiments reveal that using an RNN to compute word representations is consistently more effective than the other methods tested across a sample of eight languages with different typology and varying numbers of byte-pair tokens per word.
|
|
| 8. |
- Ek, Adam, 1990-, et al.
(author)
-
Distinguishing Narration and Speech in Prose Fiction Dialogues
- 2019
-
In: Proceedings of the Digital Humanities in the Nordic Countries 4th Conference. - : CEUR-WS.org. ; , s. 124-132
-
Conference paper (peer-reviewed)abstract
- This paper presents a supervised method for a novel task, namely, detecting elements of narration in passages of dialogue in prose fiction. The method achieves an F1-score of 80.8%, exceeding the best baseline by almost 33 percentage points. The purpose of the method is to enable a more fine-grained analysis of fictional dialogue than has previously been possible, and to provide a component for the further analysis of narrative structure in general.
|
|
| 9. |
- Ek, Adam, 1990, et al.
(author)
-
How does Punctuation Affect Neural Models in Natural Language Inference
- 2020
-
In: Proceedings of the Probability and Meaning Conference (PaM 2020), October 14-15, 2020, s. 109-116. - : Association for Computational Linguistics. - 0891-2017.
-
Conference paper (peer-reviewed)abstract
- Natural Language Inference models have reached almost human-level performance but their generalisation capabilities have not been yet fully characterized. In particular, sensitivity to small changes in the data is a current area of investigation. In this paper, we focus on the effect of punctuation on such models. Our findings can be broadly summarized as follows: (1) irrelevant changes in punctuation are correctly ignored by the recent transformer models (DistilBERT) while older RNN-based models were sensitive to them. (2) All models, both transformers and RNN-based models, are incapable of taking into account small relevant changes in the punctuation.
|
|
| 10. |
- Ek, Adam, 1990, et al.
(author)
-
How much of enhanced UD is contained in UD?
- 2020
-
In: Proceedings of the 16th International Conference on Parsing Technologies and the IWPT 2020 Shared Task on Parsing into Enhanced Universal Dependencies, July 9, 2020 / Gosse Bouma, Yuji Matsumoto, Stephan Oepen, Kenji Sagae, Djamé Seddah, Weiwei Sun, Anders Søgaard, Reut Tsarfaty, Dan Zeman (Editors). - Stroudsburg, PA : Association for Computational Linguistics. - 9781952148118
-
Conference paper (peer-reviewed)abstract
- In this paper, we present the submission of team CLASP to the IWPT 2020 Shared Task on parsing enhanced universal dependencies (Bouma, 2020). We develop a tree-to-graph transformation algorithm based on dependency patterns. This algorithm can transform gold UD trees to EUD graphs with an ELAS score of 81.55 and a EULAS score of 96.70. These results show that much of the information needed to construct EUD graphs from UD trees are present in the UD trees. Coupled with a standard UD parser, the method applies to the official test data and yields and ELAS score of 67.85 and a EULAS score is 80.18.
|
|
| 11. |
- Ek, Adam, 1990-, et al.
(author)
-
Identifying Speakers and Addressees in Dialogues Extracted from Literary Fiction
- 2018
-
In: Proceedings of the Eleventh International Conference on Language Resources and Evaluation (LREC-2018). - : European Language Resources Association. - 9791095546009 ; , s. 817-824
-
Conference paper (peer-reviewed)abstract
- This paper describes an approach to identifying speakers and addressees in dialogues extracted from literary fiction, along with a dataset annotated for speaker and addressee. The overall purpose of this is to provide annotation of dialogue interaction between characters in literary corpora in order to allow for enriched search facilities and construction of social networks from the corpora. To predict speakers and addressees in a dialogue, we use a sequence labeling approach applied to a given set of characters. We use features relating to the current dialogue, the preceding narrative, and the complete preceding context. The results indicate that even with a small amount of training data, it is possible to build a fairly accurate classifier for speaker and addressee identification across different authors, though the identification of addressees is the more difficult task.
|
|
| 12. |
- Ek, Adam, 1990, et al.
(author)
-
Language Modeling with Syntactic and Semantic Representation for Sentence Acceptability Predictions
- 2019
-
In: Proceedings of the 22nd Nordic Conference on Computational Linguistics, 30 September – 2 October, 2019, Turku, Finland / Mareike Hartmann, Barbara Plank (Editors). - University of Linköping : Linköping University Electronic Press. - 1650-3686 .- 1650-3740. - 9789179299958
-
Conference paper (peer-reviewed)abstract
- In this paper, we investigate the effect of enhancing lexical embeddings in LSTM language models (LM) with syntactic and semantic representations. We evaluate the language models using perplexity, and we evaluate the performance of the models on the task of predicting human sentence acceptability judgments. We train LSTM language models on sentences automatically annotated with universal syntactic dependency roles (Nivre, 2016), dependency depth and universal semantic tags (Abzianidze et al., 2017) to predict sentence acceptability judgments. Our experiments indicate that syntactic tags lower perplexity, while semantic tags increase it. Our experiments also show that neither syntactic nor semantic tags improve the performance of LSTM language models on the task of predicting sentence acceptability judgments.
|
|
| 13. |
- Ek, Adam, 1990
(author)
-
Studies in Language Structure using Deep Learning
- 2023
-
Doctoral thesis (other academic/artistic)abstract
- This thesis deals with the discovery, prediction, and utilization of structural patterns in language using deep learning techniques. The thesis is divided into two sections. The first section gives an introduction to the tools used and the structures in language we are interested in. The second part presents five papers addressing the research questions. The first three papers deals with discovering and predicting patterns. In the first paper, we explore methods of composing word embeddings to predict morphological features. The second paper deals with predicting the depths of nested structures. The remaining three papers deal with using structures in language to make semantic predictions. The third paper explores using dependency trees to predict semantic predicate-argument structures using a rule-based system. The fourth paper explores modeling linguistic acceptability using syntactic and semantic labels. The fifth paper deals with exploring how punctuation affects natural language inference.
|
|
| 14. |
- Ek, Adam, 1990, et al.
(author)
-
Synthetic Propaganda Embeddings to Train a Linear Projection
- 2019
-
In: Proceedings of The 2nd Workshop on NLP for Internet Freedom : Censorship, Disinformation, and Propaganda, November 4, 2019, Hong Kong / Anna Feldman, Giovanni Da San Martino, Alberto Barrón-Cedeño, Chris Brew, Chris Leberknight, Preslav Nakov (Editors). - Stroudsburg, PA : Association for Computational Linguistics. - 9781950737895
-
Conference paper (peer-reviewed)abstract
- This paper presents a method of detecting fine-grained categories of propaganda in text. Given a sentence, our method aims to identify a span of words and predict the type of propaganda used. To detect propaganda, we explore a method for extracting features of propaganda from contextualized embeddings without fine-tuning the large parameters of the base model. We show that by generating synthetic embeddings we can train a linear function with ReLU activation to extract useful labeled embeddings from an embedding space generated by a general-purpose language model. We also introduce an inference technique to detect continuous spans in sequences of propaganda tokens in sentences. A result of the ensemble model is submitted to the first shared task in fine-grained propaganda detection at NLP4IF as Team Stalin. In this paper, we provide additional analysis regarding our method of detecting spans of propaganda with synthetically generated representations.
|
|
| 15. |
- Ek, Adam, 1990, et al.
(author)
-
Training Strategies for Neural Multilingual Morphological Inflection
- 2021
-
In: Proceedings of the 18th SIGMORPHON Workshop on Computational Research in Phonetics, Phonology, and Morphology. August 2021, Online, pp. 260–267. - : Special Interest Group on Computational Morphology and Phonology.
-
Conference paper (peer-reviewed)abstract
- This paper presents the submission of team GUCLASP to SIGMORPHON 2021 Shared Task on Generalization in Morphological Inflection Generation. We develop a multilingual model for Morphological Inflection and primarily focus on improving the model by us-ing various training strategies to improve accuracy and generalization across languages.
|
|
| 16. |
- Ek, Adam, 1990, et al.
(author)
-
Vector Norms as an Approximation of Syntactic Complexity
- 2023
-
In: Proceedings of the Second Workshop on Resources and Representations for Under-Resourced Languages and Domains (Resourceful-2023), May 22, 2023 Tórshavn, Faroe Islands / editors: Nikolai Ilinykh, Felix Morger, Dana Dannélls, Simon Dobnik, Beáta Megyesi, Joakim Nivre. - Stroudsburg, PA : Association for Computational Linguistics. - 9781959429739
-
Conference paper (peer-reviewed)abstract
- Internal representations in transformer models can encode useful linguistic knowledge about syntax. Such knowledge could help optimise the data annotation process. However, identifying and extracting such representations from big language models is challenging. In this paper we evaluate two multilingual transformers for the presence of knowledge about the syntactic complexity of sentences and examine different vector norms. We provide a fine-grained evaluation of different norms in different layers and for different languages. Our results suggest that no single part in the models would be the primary source for the knowledge of syntactic complexity. But some norms show a higher degree of sensitivity to syntactic complexity, depending on the language and model used.
|
|
| 17. |
- Piementel, Tiago, et al.
(author)
-
SIGMORPHON 2021 Shared Task on Morphological Reinflection: Generalization Across Languages
- 2021
-
In: Proceedings of the 18th SIGMORPHON Workshop on Computational Research in Phonetics, Phonology, and Morphology. August 2021, Online, pp. 229–259. - : Special Interest Group on Computational Morphology and Phonology.
-
Conference paper (peer-reviewed)abstract
- This year’s iteration of the SIGMORPHON Shared Task on morphological reinflection focuses on typological diversity and cross- lingual variation of morphosyntactic features. In terms of the task, we enrich UniMorph with new data for 32 languages from 13 language families, with most of them be- ing under-resourced: Kunwinjku, Classical Syriac, Arabic (Modern Standard, Egyptian, Gulf), Hebrew, Amharic, Aymara, Magahi, Braj, Kurdish (Central, Northern, Southern), Polish, Karelian, Livvi, Ludic, Veps, Võro, Evenki, Xibe, Tuvan, Sakha, Turkish, In- donesian, Kodi, Seneca, Asháninka, Yanesha, Chukchi, Itelmen, Eibela. We evaluate six systems on the new data and conduct an extensive error analysis of the systems’ predictions. Transformer-based models generally demonstrate superior performance on the majority of languages, achieving >90% accuracy on 65% of them. The languages on which systems yielded low accuracy are mainly underresourced, with a limited amount of data. Most errors made by the systems are due to allomorphy, honorificity, and form variation. In addition, we observe that systems especially struggle to inflect multiword lemmas. The systems also produce misspelled forms or end up in repetitive loops (e.g., RNN-based models). Finally, we report a large drop in systems’ performance on previously unseen lemmas.
|
|
| 18. |
|
|
| 19. |
- Wirén, Mats, 1954-, et al.
(author)
-
Annotation Guideline No. 7 (revised): Guidelines for annotation of narrative structure
- 2021
-
In: Journal of Cultural Analytics. - : CA: Journal of Cultural Analytics. - 2371-4549 .- 2371-4549. ; 6:4, s. 164-186
-
Journal article (peer-reviewed)abstract
- Analysis of narrative structure can be said to answer the question “Who tells what, and how?”. The key part of our annotation scheme is related to the “who?”, and to this end we distinguish between narration and fictional dialogue. Furthermore, with respect to the latter we keep track of turns, lines, identities of speakers and addressees, and speech-framing constructions, which provide the narrator’s cues about the circumstances of the speech. We also annotate voice, that is, whether the narrator is ever present in the story or not. Our annotation of the “what?” includes embeddings of narrative transmission levels to capture stories in stories, and embeddings of fictional dialogue to capture characters quoting other characters. Our annotation of the “how?” includes focalization, that is, the perspective from which the narrative is seen and how much information the narrator has access to.
|
|
